AI Generates Sanitised Rewilding Landscape Images
Two geographers at the University of Aberdeen recently analyzed images generated by AI chatbots (Gemini, ChatGPT, Copilot) when prompted to depict rewilded Britain. They found models typically produce similar, sanitized 'polite' landscapes—open meadows, distant hills, and tame animals—while omitting messy processes, people, and contentious species; ecological realism appeared only with highly specific, expert prompts. This bias could shape public expectations and complicate acceptance of real-world rewilding projects.
Key Points
- 1Show AI models (Gemini, ChatGPT) produce similar sanitized rewilding images dominated by idyllic meadows and tame fauna
- 2Indicate reliance on curated NGO and social media visuals causes omission of messy processes, people, and less charismatic species
- 3Warn that sanitized AI outputs can shape public expectations, making acceptance of messy real-world rewilding harder
Scoring Rationale
Relevant empirical study reveals dataset-driven sanitization of AI images, but reliance on a single research effort limits broad generalization.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
